62 A Novel Serial Dilution Protocol for the Assessment of the Value of the Advia 1800’s Hemolytic Index as a Measure of the Degree of Plasma Red Cell Contamination

Abstract

While today’s automated chemistry analyzers routinely provide hemolytic, lipemic, and icteric indices in order to assess potential interferences in patient results, the hemolytic index (HI) is by far the best indicator of the degree of contamination of serum/plasma samples with red cell constituents. In vitro hemolysis, often due to poor sample collection and handling, is a well-known source of interference error in the core laboratory. Several analytes, especially potassium and lactate dehydrogenase (LDH), are significantly affected by contamination from red cell constituents. Since the standard of care in our laboratory involves routine testing and reporting of patient samples with an HI of 1+ or greater, a significant technologist effort is required in the decision to validate and release results. This study presents a novel serial dilution protocol to assess the degree of hemolysis/contamination interference on selected analytes including potassium, LDH, and creatine phosphokinase (CPK) with triglyceride levels included as a negative control. Ten lithium heparin whole blood specimens were analyzed for baseline laboratory values on the Rapidpoint 500 analyzer. Two aliquots were made with one frozen at –70°C for 10 minutes to cause complete hemolysis, then thawed and centrifuged for three minutes at 3,100g. Serial dilutions of the supernatant hemolysate up to 1:64 were performed using the unfrozen sample as a diluent, with results of various analytes calculated on the Advia 1800 chemistry analyzer (Siemens Healthcare Diagnostics, Tarrytown, NY). Plasma-free hemoglobin was determined on each sample on the HemoCue analyzer (HemoCue America, Brea, CA). Evaluation of various analytes on all 10 samples obtained by the Advia 1800 showed a significant increase in potassium, LDH, and CPK with an increased HI, with no significant change in the triglyceride level. An example of serial dilution revealed a decreasing potassium level of 9.2 mmol/L, LDH of 1,698 U/L, and CPK of 513 U/L at 4+ hemolysis (no dilution), to 5.3 mmol/L, 187 U/L, and 340 U/L at 1+ hemolysis, respectively (1:32 dilution), compared to a baseline potassium level of 5.2 mmol/L, LDH of 126 U/L, and CPK of 332 U/L with no hemolysis. The HI of 1+ correlates with plasma hemoglobin levels of 50–100 mg/dL, with 18 mg/dL giving an HI of 2+, 350 mg/dL giving an HI of 3+, and 1,390 mg/dL giving an HI of 4+. We conclude that HI values of 1+ on the Advia 1800 are associated with a clinically insignificant 0.1–0.2 mmol/L increase in potassium on all 10 samples. Enzymes, however, showed an increase in all HI levels.